假设我有一个3列的数据框,dt, unit, sold
.我想知道的是如何创建一个名为prior_3_avg
的新专栏,顾名思义,在过go 三个星期的同一天里,平均是sold
乘unit
.例如,对于2020年5月5日的单元"1",它在4月28日、21日和14日的平均销量是多少,这是最后三个周四?
玩具样本数据:
df = pd.DataFrame({'dt':['2020-5-1','2020-5-2','2020-5-3','2020-5-4','2020-5-5','2020-5-6','2020-5-7','2020-5-8','2020-5-9','2020-5-10','2020-5-11','2020-5-12','2020-5-13','2020-5-14','2020-5-15','2020-5-16','2020-5-17','2020-5-18','2020-5-19','2020-5-20','2020-5-21','2020-5-22','2020-5-23','2020-5-24','2020-5-25','2020-5-26','2020-5-27','2020-5-28','2020-5-1','2020-5-2','2020-5-3','2020-5-4','2020-5-5','2020-5-6','2020-5-7','2020-5-8','2020-5-9','2020-5-10','2020-5-11','2020-5-12','2020-5-13','2020-5-14','2020-5-15','2020-5-16','2020-5-17','2020-5-18','2020-5-19','2020-5-20','2020-5-21','2020-5-22','2020-5-23','2020-5-24','2020-5-25','2020-5-26','2020-5-27','2020-5-28',],'unit':[1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2],'sold':[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28]})
df['dt'] = pd.to_datetime(df['dt'])
dt unit sold
0 2020-05-01 1 1
1 2020-05-02 1 2
2 2020-05-03 1 3
3 2020-05-04 1 4
4 2020-05-05 1 5
5 2020-05-06 1 6
...
我该怎么办?我看到:
这就解释了如何在列上进行分组.我想我可以做一个"一周中的某一天"专栏,但我仍然有一个相同的问题,那就是我想限制到过go 3个匹配的一周中的某一天的值,而不是所有的结果.
这可能与此有关,但这看起来更像是用于一次性分析,而不是制作一个新专栏:limit amount of rows as result of groupby Pandas